Simon Maurer

any feedback with respect to this?

@Maratyszcza okay thanks for the feedback. issued a [pull request](https://github.com/tensorflow/tensorflow/pull/47437) in the official TensorFlow repo to allow building on platform `macos_arm64` using cmake. could you give me a hint on...

[PR](https://github.com/google/XNNPACK/pull/1327) to resolve this. allows compilation for `macos_arm64`on M1 host using: > bazel build -c opt --config=macos_arm64 :end2end_bench cross compilation on other platforms (e.g. Linux or macOS_x86-64) does not work...

@lgeiger @AdamHillier @Tombana : just as a suggestion since I'm dealing with `larq.quantizers.SteSign` and `larq.quantizers.SteHeaviside` model conversions. for `larq.math.sign` and `larq.math.heaviside` to use: ``` tf.where(x >= 0., 1., -1.) ```...

@lgeiger ah I see, this is indeed good reasoning. I didn't do any performance tests yet. I implicitly assumed you are working with `tf.float32` tensors since this is the data...

@kaivu1999 that's a really good discussion you brought up and has been of interest since the release of BMXNetv1. pointing out the findings of @yanghaojin that for training you skipped...

any news on this? still seems a problem with the latest TF2.5(RC) release